You are on page 1of 5

Assessment of disease severity and prognosis

T.A. Medsger, Jr.1, S. Bombardieri2, L. Czirjak3, R. Scorza4, A. Della Rossa2,


W. Bencivelli2

1
Division of Rheumatology and Clinical ABSTRACT ables; and (3) construct validity, i.e.,
Immunology, University of Pittsburgh The Subcommittee members initially the scoring system parallels an inde-
School of Medicine, Pittsburgh, Pennsyl- agreed on the concepts of disease activ - pendently ascertained severity measu-
vania, USA; 2Rheumatology Unit, Depart-
ity, damage and severity, defining se - rement. It is also desirable to keep the
ment of Internal Medicine, University of
Pisa, Pisa, Italy; 3Department of Immuno- verity as the total effect of disease on total number of variables to a minimum
logy and Rheumatology, Hungarian Bro- organ function. It was decided to start and to retain only those variables that
thers of St. John of God and the University with the assessment of severity using could be practically and feasibly col-
of Pécs, Pécs, Hungary; 4Clinical Immuno- the Medsger’s severity scale. A revised lected in an academic medical center or
logy and Allergology Unit, University of version of this scale was constructed. office practice setting.
Milan and IRCCS, Ospedale Maggiore, The rationale for the exclusion of other Initially, disease severity in SSc may be
Milan, Italy.
variables was provided. minimal and is likely to be mostly at-
T.A. Medsger, Jr., Professor of Medicine; tributable to activity (reversible inflam-
S. Bombardieri, Professor of Rheumatolo-
Introduction mation/edema) rather than damage (ir -
gy; Laszlo Czirjak, Professor of Rheuma-
tology; Raffaella Scorza, MD, Professor of The concepts of disease severity, dam- reversible fibrosis). Later in the natural
Clinical Immunology and Allergology; age and activity are difficult to define. history of disease, activity and damage
Alessandra Della Rossa, MD; Walter Ben- There is no agreement among rheuma- may contribute more equally to severi-
civelli , Associate Professor of Medical tologists regarding the precise meaning ty. However, in late stage disease the
Physics. of these terms. For the purpose of this greatest component of severity will be
Please address correspondence to: Prof. conference, the Subcommittee on Dis- damage (fibrosis) with little or no acti-
T.A. Medsger Jr., Division of Rheumatology ease Severity defined these conditions vity (inflammation). This concept is
and Clinical Immunology, University of as follows: important since it directs our approach
Pittsburgh, School of Medicine, 3471 Fifth
1.Severity is the total effect of disease to therapeutic intervention. Anti-in-
Avenue, Suite 502, Pittsburgh, PA 15213,
USA. on organ function; it has both irre- flammatory or immunosuppressive
versible and reversible components. treatment makes sense early in disease,
Clin Exp Rheumatol 2003; 21 (Suppl. 29):
S42-S46. 2.Damage is that component of severi- whereas anti-fibrotic therapy (no agents
ty that is irreversible. currently available and effective) is
© Copyright CLINICAL AND EXPERIMENTAL
RHEUMATOLOGY 2003. 3.Activity is that component of severity more logical in late stage disease.
that is reversible; activity may result Disease prognosis refers to outcomes of
Key words: Systemic sclerosis, in no or little damage in the future or the disease process on the host and in-
disease severity, disease prognosis. be replaced completely by damage. cludes mortality (survival) and morbid-
An important limitation of studies on ity (disability, limitations). The latter
systemic sclerosis (SSc) is the lack of a may take the form of deficits in the
standardized method to determine the ability to perform activities of daily liv-
severity of disease, either in an individ- ing, and require alterations in patient
ual organ system or globall y. A disease employment, lifestyle and psychosocial
severity scale would be extremely use- adjustment to illness.
ful in assessing disease status at a given A disease severity scale for SSc recent-
time (cross sectional) and in tracking ly has been developed and internally
the evolution of disease over time (lon- tested (1). The authors identified nine
gitudinal). Such a scale would also as- organ systems and identified variables
sist in developing stratification vari- for each one which could be used for
ables and in measuring treatment effi- defining severity. They then used pro-
cacy in clinical trials. spective data collection to determine
A severity scale should have: (1) face the feasibility of certain variables and
validity (makes sense) i.e., experts ac- received feedback describing the asso-
cept the grades within each organ sys - ciation of each variable with mortality
tem as clinically credible; (2) content in an available comprehensive longitu -
validity, i.e., all variables that are consi- dinal SSc databank as a proxy for se-
dered important are included or are rep- verity. After discussion, consensus was
resented by other equally useful vari- reached on each organ system and se-

S-42
Assessment of disease severity and prognosis in SSc / T.A. Medsger, Jr. et al.

Table I. Preliminary SSc severity scale (from reference 1)

Organ system 0 (normal) 1 (mild) 2 (moderate) 3 (severe) 4 (endstage)

1. General Normal Wt loss 5.0-9.9 kg; Wt loss 10.0-14.9 kg; Wt loss 15.0-19.9 kg; Wt loss 20+ kg;
PCV 33.0-36.9% PCV 29.0-32.9% PCV 25.0-28.9% PCV < 25.0%
2. Peripheral Normal Raynaud’s requiring Digital pitting scars Digital tip ulcerations Digital gangrene
vascular vasodilators
3. Skin TSS 0 TSS1-1-14 TSS 15-29 TSS 30-39 TSS 40+
4. Joint/tendon FTP 0-0.09 cm FTP 1.0-1.9 cm FTP 2.0-3.9 cm FTP 4.0-4.9 cm FTP 5.0+ cm
5. Muscle No proximal Proximal weakness, Proximal weakness, Proximal weakness, Proximal weakness,
weakness mild moderate severe severe; ambulation
aids required
6. GI tract Normal Distal esophageal Distal esophageal Malabsorption Hyperalimentation
hypoperistalsis; aperistalsis; anti- syndrome; required
small bowel series biotics required for episodes of pseudo-
abnormal bacterial overgrowth obstruction
7. Lung Normal DLCO 70-80%; DLCO 50-69%; DLCO < 50%; FVC Oxygen required
FVC 70-80%; basilar rales; FVC 50-69%; mild < 50%; moderate-severe
fibrosis on radiograph pulmonary hypertension pulmonary hypertension
8. Heart Normal EKG conduction defect; Arrhythmia; hypertension CHF; arrhythmia
LVEF 45-49% RVE plus LVE; LVEF < 40% requiring Rx
LVEF 40-44%
9. Kidney Normal Serum creatinine Serum creatinine Serum creatinine Dialysis required
1.3-1.6 mg/dl; 1.7-2.9 mg/dl; 3.0+ mg/dl
urine proteine 2+ urine proteine 3-4+

PCV: packed cell volume (hematocrit); TSS: total skin score; FTP: finger-to-palm distance in flection; DLCO: diffusing capacity for carbon monoxide %
predicted; FVC: forced vital capacity, % predicted; EKG: electrocardiogram; LVEF: left ventricular ejection fraction; CHF: congestive heart fauilure; RVE:
right ventricular enlargement; LVE: left ventricular enlargement.

verity scales were developed from 0 Candidate variables 3. Skin System. It was recommended
(no documented involvement) to 4 The Subcommittee elected to use the that the modified Rodnan skin thick-
(endstage disease) for each organ sys- Medsger et al. severity scale (1) as the ness scoring system (10) be retained
tem (Table I). basic document for discussion. For without change.
Thus far, only one article has been pub- each organ system, the subcommittee 4. Joint/Tendon Systems. Joint involve-
lished in which the Medsger et al. se- thoroughly discussed the published 0-4 ment is typically described as consist-
verity scale has been utilized. One hun- scale items and made recommenda- ing of joint pain on motion, tenderness
dred Swedish SSc patients followed tions for approval or modification. or swelling (synovitis or effusion).
over 14 years had organ system in- These features are amenable to descrip-
volvement assessed using some (but Discussion tion in diseases which affect only the
not all) of the original scale items (2). Identification of core set variables articular structures, such as rheumatoid
In this report, as expected, a high sever- 1.General System. Calculation of arthritis. However, in SSc, where skin,
ity score was shown to predict reduced weight loss in Kg should begin with the subcutaneous tissue, tendon sheaths
survival. The primary determinants of patient’s baseline weight immediately and tendons themselves are all affected,
poor survival were extensive skin prior to the onset of SSc. A percentage and by both inflammation and fibrosis,
thickening, ECG changes and reduced of total body weight loss from baseline the use of tender/swollen joint counts is
lung and renal function. should be used rather than absolute less reliable and less reproducible. Pal-
The original intent of the severity scale weight loss. 0 (normal) = < 5%; (1) pable tendon friction rubs (11) are con-
was that it should include the concept mild = 5-10%; (2) moderate = 10-15%; sidered evidence of activity (fibrous
of improvement. Unlike the damage (3) severe = 15-20%; and (4) endstage tenosynovitis, tendinitis) rather than
scale in systemic lupus erythematosus = > 20%. Hemoglobin is an acceptable damage.
(3), where damage continues to accu- alternative to packed cell volume (PCV It is recognized that joint contractures
mulate without reversal, SSc severity or hematocrit),as follows: 0 (normal) = can represent the end result of patho-
can impr ove. This has been shown for 12.3 Gm/dl or greater; 1 (mild) = 11.0- logic processes in all of the above tis -
skin thickening (4, 5), pulmonary func- 12.3; (2) moderate = 9.7-11.0; 3 (severe) sues plus skeletal muscles. Considering
tion (6, 7) and renal disease (8, 9). In = 8.3-9.7; and (4) endstage=<8.3 Gm/dl. the nearly uniform loss of motion of the
the Swedish study, organ dysfunction 2. Peripheral Vascular System. The pu- finger joints, finger contracture was felt
accumulated in the first five years of di- blished scale was considered adequate to be the best candidate variable to in-
sease and remained stable/unchanged to describe the spectrum of digital vas- clude. The published methods of the
thereafter (2). cular ischemia. 3rd fingertip to distal palmar crease

S-43
Assessment of disease severity and prognosis in SSc / T.A. Medsger, Jr. et al.

Table II. Revised preliminary SSc severity scale

Organ system 0 (normal) 1 (mild) 2 (moderate) 3 (severe) 4 (endstage)

1. General Wt loss < 5%; Wt loss 5.0-9.9%; Wt loss 10.0-14.9%; Wt loss 15.0-19.9%; Wt loss 20+ %;
PCV 37.0%+; PCV 33.0-36.9% PCV 29.0-32.9% PCV 25.0-28.9% PCV < 25.0%
Hb 12.3+ Gm/dl Hb 11.0-12.2 Gm/dl Hb 9.7-10.9 Gm/dl Hb 8.3-9.6 Gm/dl Hb < 8.3 Gm/dl
2. Peripheral No Raynaud’s; Raynaud’s Raynaud’s requiring Digital pitting scars Digital tip ulcerations Digital gangrene
vascular not requiring vasodilators vasodilators
3. Skin TSS 0 TSS 1-14 TSS 15-29 TSS 30-39 TSS 40+
4. Joint/tendon FTP 0-0.09 cm FTP 1.0-1.9 cm FTP 2.0-3.9 cm FTP 4.0-4.9 cm FTP 5.0+ cm
5. Muscle Normal proximal Proximal weakness, Proximal weakness, Proximal weakness, Ambulation aids
muscle strength mild moderate severe required
6. GI tract Normal esophagram; Distal esophageal Antibiotics required Malabsorption Hyperalimentation
normal small bowel series hypoperistalsis; small for bacterial over- syndrome; episodes required
bowel series abnormal growth of pseudo-obstruction
7. Lung DLCO 80+%; DLCO 70-79%; FVC DLCO 50-69%; DLCO < 50%; Oxygen required
FVC 80+%; 70-79%; basilar rales; FVC 50-69%; FVC < 50%;
No fibrosis on radiograph; Fibrosis on radiograph sPAP 50-64 mmHg sPAP 65+ mmHg
sPAP < 35 mmHg sPAP 35-49 mmHg
8. Heart EKG normal; EKG conduction defect; EKG arrhythmia; EKG arrhythmia requir- CHF;
LVEF 50+% LVEF 45-49% LVEF 40-44% ing Rx; LVEF 30-40% LVEF < 30%
9. Kidney No Hx SRC with serum Hx SRC with serum Hx SRC with serum Hx SRC with serum Hx SRC with serum
creatinine < 1.3 mg/dl creatinine < 1.5 mg/dl creatinine 1.5-2.4 mg/dl creatinine 2.5-5.0 mg/dl creatinine > 5.0 mg/dl
or dialysis required

Wt: weight; PCV: packed cell volume (hematocrit); Hb:hemoglobin; TSS:total skin thickness score; FTP: fingertip-to-palm distance in flexion; DLCO: dif-
fusing capacity for carbon monoxide, % predicted; FVC: forced vital capacity, % predicted; sPAP:estimated pulmonary artery systolic pressure by Doppler
echo; EKG:electrocardiogram; LVEF:left ventricular ejection fraction; Rx:treatment; CHF:congestive heart failure; Hx:history of; SRC:scleroderma renal
crisis.
N.B. If two items are included for a severity grade, only one is required for the patient to be scored as having that severity level.

(FTP) measurement (12) and hand- administered by the examining phy- 7. Lung System. The lung has two pri-
spread (13) were considered. Both have sician, as currently recommended by mary types of involvement, i.e. intersti-
limitations. The FTP however does the severity scale authors. However, tial inflammation/fibrosis and pulmo-
have a normal value of 0.0 cm, even if this method will require detailed in- nary vascular disease. Each should be
the measurement technique is not stan- structions for the examiner and valida- accommodated by the severity scale.
dardized and the variability between tion before it is acceptable as a measure Subcommittee members agreed that the
examiners is large. Also, fingers which of skeletal muscle dysfunction. spectrum of pulmonary function test
are so severely contracted that they are 6. Gastrointestinal System. The sub- results, particularly the FVC percent
“fixed” in exaggerated flexion have a committee felt that a radiographic predicted (restrictive disease) and
small FTP distance, which suggests method was preferable, supplemented DLCO percent predicted (vascular dis-
only minimal impairment. Such hands by “clinical judgement”, with the eso- ease) should be included. The DLCO
are actually very poor from a function- phagus and small intestine being the should be corrected for alveolar volu-
al standpoint since they have virtually primary sites for evaluation. Small inte- me (DLCO/VA).
no motion. The subcommittee mem- stinal radiographic abnormalities alone 8. Heart System. It was generally agre-
bers felt that a more precise/reliable need to be distinguished from the clini- ed that the most important abnormali-
measure of finger contracture would be cal disorders which can result, i.e. bac- ties include those resulting from dys-
most desir able, and thus issues a chal- terial overgrowth (hydrogen bre at h function of the conduction system and
lenge for clinical investigators to devel- test) with diarrhea and the more severe the left ventricular myocardium. Thus
op such a measure. In the future, anoth- manifestations of episodes of function- routine electrocardiogram and, if avail-
er alternative would be a patient-com- al small intestinal obstruction (pseudo- able, Holter monitor tests are needed to
pleted hand function questionnaire or obstruction) and frank malabsorption detect conduction system abnormalities
practical test. However, for the present syndrome. The subcommittee recom- and arrhythmias and echocardiography
we recommend retaining the FTP mea- mended continuing to use the pub- to quantitate LV contractile function.
surement as described in the severity lished severity scale for the gastroin- The published list of mild, moderate,
scale publication, without any changes. testinal system, with better descriptions severe and endstage manifestations
5. Skeletal Muscle System. The sub- provided for the examiner to facilitate might be modified in the future.
committee agreed to the use of the pro- proper classification of the individual 9. Kidney System. For renal disease, the
ximal muscle strength grading system patient. Subcommittee felt that reorganization

S-44
Assessment of disease severity and prognosis in SSc / T.A. Medsger, Jr. et al.

of the severity scale was necessary. 3. Skin. Ultrasound methods were con- 9. Kidney. The evaluation of the glome-
Because scleroderma renal crisis sidered but the equipment is expensive rular filtration rate could be carried out
(SRC) is the most frequent and domi- and not uniformly available and the by an isotopic technique, if an inexpen-
nant renal manifestation, its presence results are operator-dependent and not sive and uniformly available procedure
or absence was considered to be the yet standardized. were to be established. This methodo-
most important distinction between no 4. Joint/tendon. Hand spread was con- logy is not feasible in many centers.
involvement (no history of SRC) and sidered. However, it has no normal val-
the varying degrees of renal disease, ue, and thus a handspread distance can- Mortality
grade 1-4 (history of SRC). It was re- not be related to a “baseline” measure- Reports on survival should consider
commended that these latter grades be ment to determine the degree of abnor- methods of calculating survival. The
associated with specific levels of serum mality attributable to the disease. The death date is the endpoint for all such
creatinine or a simple method for esti- detection of joint contractures was be- calculations. The starting date can be
mation of creatinine clearance such as lieved by the Subcommittee to be the any one of three dates: (1) date of dis-
the Cockroft index (14). For example, best available, but not ideal, measure to ease onset (first symptom or finding at-
(mild) could be a history of SRC and use. tributed to SSc) as judged by the pa-
current serum creatinine < 1.5 mg/dl or 5. Muscle. No other variable was con- tient or the physician; (2) date of the
creatinine clearance > 100 cc/min; 2 sidered. first physician diagnosis of SSc; or (3)
(moderate) = 1.5-2.4 or 60-99; 3 (se- 6. Gastrointestinal. To assess esopha- date of first enrollment in a research
vere) = 2.5-5.0 or 30-79; and 4 (end- geal involvement, a quantitative mano- study. Cumulative survival from the
stage) = dialysis, 5.0 mg/dl or greater metric method also would be helpful, starting date should be calculated using
or clearance < 30 cc/minute. but is not widely available. In addition, a standard method and taking into
definitions would be necessary for “hy- account patients lost to follow-up, such
Rationale for the exclusion of other poperistalsis” and “mild to moderate as the Kaplan-Meier method. All re-
variables manometric abnormalities” so that they ports should include 5- and 10-year cu-
1. General system. Serum albumin was could be distinguished from “aperistal- mulative survival rates so that publish-
considered as a possible variable to sis” and “severe manometric abnorma- ed studies can be directly compared
describe general organ dysfunction, as lities”. The SSc HAQ GI VAS (15) was with one another.
follows: 0 (normal) = > 3.5 gm/dl; 1 also discussed as a candidate variable, The cause of death should be recorded
(mild) = 3.2-3.5; 2 (moderate) = 2.9- but it still needs to be validated. as accurately as possible, using the me-
3.2; (3) severe = 2.6-2.9; and 4 (endsd- 7. Lung. Restrictive disease should be thod described by Geirsson et al. (2)
tage) = < 2.6. It was not recommended further characterized by HRCT (1.0 which includes the categories of: (1)
as a substitute for either weight loss or mm slices) and vascular disease by definitely SSc-related (due to organ
hemoglobin. In addition, there was dis- echo Doppler or right heart catheteriza- failure [specific organ]); (2) probably
cussion concerning patient self-evalua- tion. An oxygen desaturation test could SSc-related (due to a complication
tion e.g. the HAQ disability index and be also useful. However, these techni - caused or aggravated by SSc associated
physician global assessment (VAS) as ques are not feasible everywhere. A with organ injury or treatment); (3)
reflecting the General System (15), but precise rearrangement of the published possibly SSc-related (due to a manifes-
there was concern that the responses severity scale va ri ables was not at- tation reported to have increased preva-
might reflect the most serious organ tempted during this conference. The 6- lence in SSc, such as malignancy or
system involvement. For example, the minute walk distance (16) could also be suicide); and (4) unrelated to SSc, its
patient with moderate to severe pul- considered but its interpretation is diffi- organ involvement or treatment.
monary fibrosis would answer this cult in patients with SSc who often
question based on the particular limita- have musculoskeletal factors limiting Morbidity
tions imposed by his/her lung disease. their ambulation. An SSc HAQ VAS for Important outcomes in addition to sur-
2. Peripheral vascular. For the assess- lung disease could also be included but vival are morbidity and disability relat-
ment of severity of Raynaud’s pheno- dyspnea in SSc may sometimes be at- ed to SSc involvement and treatment
menon, an alternative could be the tributable to problems with other organ complications. We recommend the
visual analog scale (VAS) from the systems and is thus not lung-specific. most widely used instrument, the disa-
Scleroderma Health Assessment Ques- 8. Heart. Pericardial disease was not bility index (DI) of the HAQ, a validat-
tionnaire (SHAQ) (15) with “mild” discussed, but could also be incorporat- ed and reliable instrument (17). HAQ
defined as >50% on the VAS. However, ed in terms of the variables small, mod- DI scores correlate well with the extent
this approach is not commonly used. In erate or large effusion and pericardial of skin thickening, loss of first closure,
addition, quantitation of capillary tamponade. Stress echocardiogram and proximal muscle weakness and tendon
microscopic abnormalities was consid- thallium perfusion studies or MUGA friction rubs. The SHAQ “global” VAS
ered to be a variable with relevance to scans were discussed as candidate mea- is an alternative (15), and a variety of
microvascular disease, but no scale for sures, but not incorporated as definite other validated instruments could be
its use has been developed to date. recommendations at this time. used. Several instruments have recently

S-45
Assessment of disease severity and prognosis in SSc / T.A. Medsger, Jr. et al.

been developed to measure hand dys- sis (scleroderma) with interstitial lung dis- health assessment questionnaire. A rt h ri t i s
ease. J.Rheumatol. 1993; 20: 838-44. Rheum 1999; 42: 2372-80.
function, but require observation of the
7. STEEN VD , OWENS G, REDMOND C, et al.: 14. COCKCROFT DW, GULT MH: Prediction of
patient and scoring by a trained profes- The effect of D-penicillamine on pulmonary creatinine clearance from serum creatinine.
sional (18-20). findings in progressive systemic sclerosis. Nephron 1976; 16: 31-41.
Arthritis Rheum 1985; 28: 882-8. 15. STEEN VD , MEDSGER TA, JR.: The value of
References 8. STEEN VD, COSTANTINO JP, SHAPIRO AP, et health assessment questionnaire and special
1. MEDSGER TA, JR., SILMAN AJ, STEHEN VD, et al.: Outcome of renal crisis in systemic scle- patient generated scales to demonstrate
al.: A disease severity scale for systemic scle- rosis: relation to availability of angiotensin change in patients with systemic sclerosis
rosis: development and testing. J Rheumatol convertine enzyme (ACE) inhibition. Ann over time. Arthritis Rheum 1997; 40: 1984-
1999; 26: 2159-67. Intern Med 1990; 113:352-57. 91.
2. GEIRSSON AJ, WOLLHEIM FA, AKESSON A: 9. STEEN VD, MEDSGER TA, JR.:Long term out- 16. BADESCH DB , TAPSON VF, MCGOON MD , et
Disease severity of 100 patients with sys- come of scleroderma renal crisis:Ann. Intern. al.: Continuous intravenous epoprostenol for
temic sclerosis over a period of 14 years: Med. 2000; 133: 600-3 pulmonary hypertension due to scleroderma
using a modified Medsger scale. Ann Rheum 10. CLEMENTS PJ, LACHENBRUCH PA, SEIBOLD spectrum of disease. Ann Intern Med 2000;
Dis 2001; 60: 1117-22. JR, et al.: Inter and intraobserver variability of 132: 425-34.
3. GLADMAN D , GINZLER E, GOLDSMITH C, et total skin thickness score (modified Rodnan 17. FRIES JF, SPITZ P, KRAINES RG , et al.: Mea-
al.: The development and initial validation of TSS) in Systemic Sclerosis. Rheumatol 1995; surement of patients outcome in arthritis.
the systemic lupus international collaborating 22:1281-5. Arthritis Rheum 1980; 23: 137-45.
clinics / American College of Rheumatology 11. STEEN VD, MEDSGER TA, JR.: The palpable 18. SANDQVIST G, EKLUND M: Hand mobility in
damage index for systemic lupus erythemato- tendon friction rub: an important physical scleroderma (HAMIS) Test: The reliability of
sus. Arthritis Rheum 1996; 39: 363-9. examination finding in patients with systemic a novel hand function test. Arthritis Care and
4. STEHEN VD, MEDSGER TA, JR.: Improvement sclerosis. Arthritis Rheum 1997; 40: 1146- Research 2000; 13: 369-74.
in skin thic kening in systemic sclerosis asso- 151. 19. SANDQVIST G, EKLUND M: Validity of
ciated with improved survival. A rt h ri t i s 12. CLEMENTS PJ, FURST DE, WONG WK, et al.: HAMIS:A test of hand mobility in scleroder-
Rheum 2001; 44: 2828-35. High-dose versus low-dose D-penicillamine ma. Arthritis Care and Research 2000; 13:
5. BLACK C, DIEPPE PK, HUSKISSON T: Pro- in early diffuse systemic sclerosis:analisys of 382-7.
gressive systemic sclerosis. Ann Rheum Dis two-year, double-blind, randomized, con- 20. POOLE JL,GALLEGOS M,O’LINE S: Reliabili-
1986; 45:384-88. trolled clinical trial. Arthritis Rheum 1999; ty and validity of the arthritis hand function
6. SILVER RM, WARRICK JH, KINSELLA MB , et 42: 1194-203. test in adults with Systemic Sclerosis (sclero-
al.: Cyclophosphamide and low dose predni- 13. CLEMENTS PJ, WONG WK, NURWITZ EL et derma). Arthritis Care and Research 2000;
sone therapy in patients with systemic sclero- al.: Correlates of the disability index of the 13: 69-73.

S-46

You might also like